Professional Documents
Culture Documents
J58E
ttJi\iversity of Illinois)
Abstract
In this paper i present dn analysts fJ/ What fe1iJii,tns ofit trade born in the seventeenth cen-
tury as a consequence 0/ Basque fishing trips in Me North Atlantic.
Since the data have show a linguistic syifiiii in an early stage ofdevelopment and the lan-
gUages in contact are easily identifiable, "whichever characteristics £n the contact language
will be traceable to one ofthe initial languages or to unilvifsal tendencies 1• For this purpose, an analysis
ofthe main typological features ofboth Basque and Icelandic is also provided. * '
1. Introduction
The only remaining vestige that we have of what appears to have been used as a
means of communication between Basque whale hunters and Icelandic traders in the
seventeenth century is constituted by a few sentences, together with a word list. The
existence of this p'idgin seems to have gone virfually unnoticed. Hancock (1977)
does not mention it in his "Repertoire of Pidgin and Creole Languages" although he
mentions an extinct Icelandic French Pidgin.
My source of information is Nicolaas Deen's Glossaria duo Vasco-lslandica (Ams-
terdam 1937, reprinted in this volume), which' is a commentary and translation of
two manuscripts written in Iceland in t4e'seventeenth century. Since the contact sit-
uation was interrupted in the 'first part of the eighteenth century and was of inter-
mittent nature, the contact pidgin probably never developed much further than the
stage recorded in the manuscripts. The' first manuscript lacks interest for our pur-
poses since it is just a list of lexical items seemingly elicited from the Basque sailors
,by pointing at different objects and vaguely organized according to semantic fields.
Most of the second manuscript contains information. Only at the end of
the second manuscript do' we find- something of a different nature. Here the listed
* This paper was originally published in the Journal ofBasque Studies in America, vol 5 (1984),41-59.
(1) One must keep in mind, however, that at least some Basque sailors may have had a certain familiarity
with Gascon, French and Spanish. The presence in the pidgin items from other languages, such as
English, does not imply, on the other hand, any knowledge of the structure of these languages on the part of
the users of the pidgin. It is possible that a greater influence can be imputed to Icelandic French Pidgin, in
case this developed earlier, which I have not been able to confirm, as well as to other pidgins.
[ASJU, XXV - 2, 1991, 427-437]
428 JOSE IGNACIO HUALDE
items include conjunctions and other parts of speech hardly obtainable by direct el-
icitation. Now we find not only Basque words or French and Spanish words possibly
borrowed into the language of the Basque seamen, but also words from other or-
igins, like clinke, translated as 4Iittle', possibly from some Germanic language; morta,
from Spanish, translated as 4to kill' and certainly never incorporated into Basque;
even what seems to be an Icelandic word with ,a Basque article, kichera (from ky-
ker+a, according to Deen)2. More interestingly, some sentences are included which
certainly are not Basque. What we have at the end of the second manuscript is a
sample of a real Hmixed language". It is to the study of this contact language that
this paper will be dedicated.
The fact that the materials from the trade pidgin are not separated from what
seems to be a pure Basque vocabulary list may indicate that in the mind of the
Icelandic scribe everything was Basque. This is in accordance with what Hall (1966:
12) reports from Russenorsk:
Each side was firmly convinced it was speaking the other's language and so the
slogan sentence' of Russonorsk {sic] was /moja p:J rvoja/ 441 {will speak] in your {lan-
guage]."
Even if the lexicon of our pidgin is mostly Basque, the structure of the sentences
does not resemble that of Basque. This is hardly surprising. It has been noticed that
the grammar of a pidgin or a creole lexically based on a language A may be quite
similar to that of another pidgin or creole spoken in a different part of the world,
but diverge widely from that of language A. DeCamp (1977) points out that French
based creoles spoken in the Indian Ocean as well as in the Caribbean are mutually
intelligible but unintelligible with French.
Certain surprising similarities amo,ng pidgins and creoles spoken in places dis-
tant from each other and with different linguistic components in their formation
have led some linguistic to formulate a monogenetic theory that would trace many
pidgins and creoles to a common ancestor: A Portuguese based pidgin or even to the
Lingua Franca or Sabir spoken in the Mediterranean Basin the Middle Ages
(Whinnom, 1977).
An opposite view would see the of a pidgin as a result of simplification
strategies that would lead the participant's to reduce the grammar of a language to a
rudimentary manner (Ferguson and DeBose, 1977). Those points in which the basic
grammar of the languages in contact clearly diverged would be decided by bor-
rowing the rule from either language or rather by relying on Universal Grammar, in
the opinion of Givon (1979).
Bickerton (1983) on the other' hand, concludes from the study of Hawaiian En-
glish Pidgin and Creole that pidgins do not have a stable grammar; this being ac-
commodated to those of the native languages of different speakers. In those circums-
(2) Deen does not offer a translation for'the supposedly Icelandic word kyker. The closest in Standard Ice-
landic seems to be kikir 'telescope: Since IPB kichera is translated as syndu mier 'show me; Deen's interpreta-
tion is possibly erroneous. This is not a noun a Basque ankle but an IPB verb, which, like the other IPB
verbs, ends also in -a (cf: presenta, "co give; travala «to do; trucka 'to buy'). It is very likely that kichera is rel-
ated co Icelandic kikja 'look through the glass; and Dutch kijken "to look.'
GLOSSARIA VASCO ISLAND/CA: ICELANDIC BASQUE PIDGIN 429
tances, the first generation of creole speakers must fix the structure of the language
according to their innate grammar. This is what produces great grammatical simil-
arities among creoles. In Bickertonts view the innate human grammar would be creole
grammar and this is what children will produce in the absence of a unified input
that forces them to adapt their innate grammar to that of the community in which
they were born.
In any case, in spite of observation, a certain amount of grammatical
coherence is to be expected in any pidgin. At 'the very least, the syntax or the mor-
phology has to provide certain ways to determine basic semantic relations like those
of agent and patient. In some way (including intonation) statements, questions and
commands need also to be distinguished in the linguistic signal to make possible the
simplest linguistic interaction.
It has that for a truly viable pidgin to develop, at least three groups
of people speaking mutually unintelligible languages are needed. In the case that
concerns us here, there are only two groups: Basque and Icelandic speakers. The data
that we have show indeed a contact language of the most mdimentary kind, adequate
only for a basic commercial interaction.
The sociolinguistic aspects of the contact situation may lead us to seek for the
closest parallel in Russenorsk. In both cases we have two groups of Europeans without
a common language which temporarily but repeatedly enter into contact for commer-
cial reasons.
In our case one of the two languages, Basque, does not belong to the Indoeurop-
ean family of languages.
In the rest of this paper I will examine the structural characteristics of what I will
baptize Icelandic Basque Pidgin in relation to Basque, Icelandic and other pidgins.
For this purpose, I first need to outline the main typological characteristics of Bas-
que and Icelandic. '
2. Basque
2.1. Word order and case marking.
The order of constituents in the Basque sentence seems to be quite free. In fact,
the main constituents are ordered in terms of pragmatic functions in a way similar
to Hungarian. Whichever' constituent immediately precedes the verb is understood
as the focus of the sentence:
(1) a. Mikel-ek txori-a ikusi du c. txori-a Mikel-ek ikusi du
FOe Foe
Mikel-ERG bird-Det seen has d. Mikel-ek ikusi du txori-a
b. txori-a ikusi du Mikel-ek Foe
FOe 'Mikel saw the bird.'
(la, b) are answers to the question 'what did Mikel see?t whereas (lc, d) answer to
'who saw the bird?'. Moreover, Mikel would be the topic in (la) and txoria 'the bird'
would be the topic in (le).
430
3. Icelandic
3.1. Word order and case marking.
Icelandic has four cases (nominative, accusative, dative, and genitive), two num-
bers, and three genders. The most frequent word order is SVO as in (14a), but other
orders are also possible (14b, c)7:
(14) a. Ari las Njalu b. las Ari Njalu
Ari-NOM read Niala-ACC c. Njalu las Ari.
The same SVO order is observed in subordinate clauses as well:
(15) peir sogdu ad Gfsli vaeri veikur -
'They said that Gisli was sick.'
In certain syntactic constructions the verb precedes i,ts subject: yes-no questions,
commands, main clauses following the subordinate clause, certain coordinated sen-
tences, and conditional clauses when ef'if' is omitted. Verb-subject inversion can
take place for stylistic purposes as well.
The verb agrees with its subject NP in person and number.
There are certain verbs which may take two accusative NP's or one accusative and
one -dative NP, but no nominative NP. In these cases the verb is marked for third
person singular agreement. Here also the NP which could be identified as "logical
subject" precedes the verb in the most normal order:
(16) mig - vantar sko (17) mig pyrstir
I-ACC need shoes-ACC I-ACC get thirsty
3.2. Noun phrase.
Within the NP the adjective precedes the noun.- The article can be free and in
initial position, like in (18a) or, more commonly suffixed to the noun like in (lSb):
(18) a. hinn godi madur b. godi madur-inn
the good man good man-the
Basque and Icelandic share the feature of having suffixed articles, obligatorily in
Basque, and optionally but preferably in Icelandic. Unlike Basque, in Icelandic an
adjective must agree in gender, case and number with its noun. There is a strong and
a weak declension for adjectives.
3.3. Other constructions.
Noun-Gen:
(19) hus Jons or (20) Hus-id hans Jons
house John-GEN house .;.the his John-GEN8
N-Rel C: Prep-NP:
(21) madur-inn sem sa mig (22) 10mb mea- mommum sinum
man-the who saw me 'Lambs with their
Comparative Adj-Standard:
(23) Jon var eldri og reyndari en eg
'John was older and more experienced than I.'
Icelandic is, therefore, consistently head-initial, the only exception being that ad-
jectives precede nouns9 •
(33) cavinit trucka for mi (n, 223) (34) for ju mala gissuna (n, 226)
eckert kaupe eg pu ert madur
'I don't" buy 'you are a bad man'
(35) Christ Maria presenta for mi balia, for mi, presenta for ju bustana (II, 224)
gefe Christur og Maria mier hval, skal jeg.gefa pier spordenn
'If Christ and Mary, give me a whale, I will give you the tail'
(36) presenta for mi berrua usnia eta bura (n, 227)
gefdu mier heita miolk og nyt: smior
'give me hot milk and new butter'
(37) ser travala for ju (11, 228)
hvad giorer pu
'what are you doing?'
5.2. Analysis of the data.
With the obvious cautions that such an exiguous representation of a linguistic
system us to take" we some observations. · _
. have a unique ,form inflecte4 for tense or aspect and with no agreement
features. The verb presenta is rendered as 'an imperative in (24), (26) and (36), present
indicative in (27), future in (28) and (35) and present subjunctive (?) in (35), in the
corresponding Icelandic translations.
Nouns are not marked for case. All nouns and adjectives end in -aJ the Basque
determiner. This even in cases where it would not appear in Basque, like after a nu-
meral (30,31, 32)'or when a noun is followed by an adjective.
Without case marking or verb agreement, the interpretation of grammatical
functions has to rely solely upon word order..The order appears to be Subject-Verb-
Indirect Object-Direct Object. The subject precedes the object in accordance with
Greenberg's (1963) first universal and what we"have seen is the normal order in Ice-
landic and the order in Basque when case marking is neutralized. In the medial posi-
tion of the verb Icelandic Basque Pidgin (IBP henceforth) takes side with Icelandic
but does not clash with what is in Basque. · .
If no NP precedes the verb. the sentence is interpreted as an imperative- (24, 25,
26, 29). As we cap see from the Icelandic glosses, in this language the second person
pronoun is suffixed to -the verb in c<?mmands 1o • In Basque, commands are formed
with the radical of the verb plus an auxiliary carrying aspect and agreement affixes.
However, this auxiliary can be omitted. (38a) and (38b) are equivalent in most con-
texts:
. (38) a. eman eza-ida-zu liburu-a b. eman liburu-a
give ASP-DAT-ERG book-the 'give me the
. Is' 2s
As a matter of fact, the,IBP sentence (29) could be an acceptable Basque sentence.
For the rest, the order 'of constituents is generally the same in IBP and in the IC,e-
landic glosses. One is (34); in which the verb precedes its subject in 'the
main clause in the Icelandic translation but not in This does not indicate any
clear divergcnc.e in word' order between IBP and Icelandic, since, pres-
umably the same Icelandic sentence could have been produced with the conjunction
ef' 'if and without inversion. "
There are two IBP sentences, (33) and (37)' where the subjec:t appears postver-
bally. In (33) the verb is preceded by cavinit, of dubious' interpretation and glossed as
ecleert 'nothing.. ' and in (37) by ser 'what' (Basque).., In both c3$es there is subject in-
version in the Icelandic glosses as well. Presumably (33) could be interpreted not
only as 'I don't buy anything' but also ambiguously 'do not buy anything for me.'
The inversion of the. subject and fronting of the question word in (37) is explain-
able not only foom the side of Icelandic that requires this transformation, but also
from the side of Basque, in which question words are necessarily the focus of the sen-
tence and must precede the verb immediately. Therefore the order in (37) would be
acceptable for speakers of both Icelandic and However, sentences involving
more than one human NP would be difflcult to interpret in IBP if the subject is not
encoded before the verb. A sentence like (39) would presumably be'ambiguous:
,(39) presenta for mi for ju
tail give I you
The interpretations would be 'I give you the tail' and 'you give me the
tail.'
E_qually ambiauous would be a sentence like (27b):
(27) a. *ser for ju for me
'whaf dQ I/yol1 give you/me?'
In (27b) 'wQrd has' been fronted immediately before the verb _fol-
, h)wing Icelandic and Basque rules) but lack of case marking and' agreement gives an
ambiguous in IBP,
A was found and is illustrated io'(27). Both question word
an.d subject precede' the verb, the subject appears in a reduced form:ju (as
compared with far jH in the' of the In this way, distance between
question'word and v@rb is somewhat reduced and the subject is also successfully eo""
coded. Presumably (21') would have an acceptable result as well; although it
deviates more sharply from Icelandic _model:
(27) c. for jll, §@f' for mill
Turning now fO @xEimples, from (34) we that the, con4irion precedes the
conclusion, in with 14th universal. Otherwise, is no
Q{- §yi)tactical jlJXtflposition of the two clauses and the
irreality of t:he first propos.ition U§ ft) Q@five the that this is a
ditional construqtion. .
(35) zero copula, against bgth and but in
with one of processe.s identified -by (1971) for simplified c,odes. Zer'Q
(11) I indicate with an asterisk (*) hypothetical seiltences and _with an arrow (i)
tical grammatical sentences. '
436 JOSE IGNACIO HUAlDE
copula is common in many languages, like Russian, Arabic and Hungarian, and in-
terestingly, in many creoles.
The order of elements within the- NP can be observed in (35) and (36). The
ordering of adjective and noun follows the Icelandic model. In addition, we can
observe in these examples that the determiner -a, which in Basque can occur only at
the end of the NP, is interpreted as part of the noun or adjective and, therefore, all
nouns and adjectives end in -a, as'we said above:
(36) a. IBP: ,berrua usnia eta berria bura
hot milk and new butter
b. Basque: esne bero-a' eta gurin berri-a
milk hot-the and butter new-the
We have not included in our sample other phrases or sentences that appear in the
text and that, when and if used by Icelandic speakers, were obviously unanalyzed
chunks, like caca hiarinsat, literally in Basque 'shit for thee' and glossed as et Pu skyt
ur rasse, whose Spanish and German translations not daring to provide, Deen
translates as ede excrementum ani.
In summary, this sample gives us the impression of a very rudimentary pidgin.
Sentences can have the basic functions of expressing .a statement, asking a question
or giving a command. Special questions are identified by wh fronting and com-
mands by subject deletion. No temporal or aspectual notions can be expressed (pro-
bably they would be expressed by means of adverbs when required). The notions of
subject and object are encoded by word order. The lexicon is drawn mainly from
Basque and, in a smaller proportion, from Spanish and French. A few lexical items
are from other origins. Curiously the first and second person singular pronouns seem
to be taken from English.
The order of constituents seems to follow the Icelandic model: SVO, Adjective
Noun, prepositions (/or). Of these, though, only the order of Adjective and Noun
could be argued to be taken from Icelandic. The order SVO is also frequent in Bas-
que and, moreover, has been characterized as the perceptually preferable order (Gi-
von 1979). The status of/or as a preposition is not transparent from the data. For ju
and/or mi perhaps were taken as monomorphemic. The only occurrence ofjor not in
pronominals is (30), in whi<:h/or could be interpreted as a stranded preposition or as
a postposition.
In commands, the omission of the subject is in accordance with the facts of Bas-
que (and the other possibly relevant languages, Spanish and French). Subject inver-
sion as in Icelandic would render sentences difficult to interpret in IBP. The omis-
sion of the copula, which cannot be traced to either language, seems to be dictated
by a principle of maximum simplicity.
Unlike other pidgins, IBP does not present a morphological marker to distin-
guish verbs from nouns (cf. Russenorsk -om, 'attached to the verb)12 or nouns from
adjectives (cf. Tok Pisin -pela, attached to the adjective).
(12) Slobin (1979: 43-46), basing his argument on the evidence of Russenorsk, claims that in the absence
of other means to establish grammatical relations than word order, some morphological means to distinguish
verbs from nouns is necessary. This marking, however, is absent in many other pidgins. Some other particles
that can serve this function of singling out the verb in the sentence are the verbal prefix i- in Tok Pisin (Alt-
hough not all verbs are preceded by i); che suffix -im attached. to transitive verbs, also in Tok Pisin, and
tense/aspect markers such as tabaltallo in Papiamento.
GLOSSARIA VASCO ISLANDICA: ICELANDIC BASQUE PIDGIN 437
The data do not show us how genitive constructions were formed. More complex
constructions, like relative clauses, perhaps did not develop. Yes-no questions are not
illustrated either; possibly they were distinguished by means of intonation only, as
seems to be general for creoles (Bickerton 1983).
References
Alvarez Enparantza, J. L., 1978, Euskal Gramatika. San Sebastian: Ediciones Vascas.
Bickerton, D., 1983, "Creole Languages", Scientific American, July.
Comrie, B., 1978, "Ergativityr' in Syntaaic Typology: Studies in the Phenomenology 0/ Language. ed. W.
P. Lehmann. Austin: Univ. of Texas Press.
- - - - , 1981, Language Universals and Linguistic Typology. Chicago: Univ. of Chicago Press.
Deen, N. G. H., 1937, Glossaria Duo Vasco-Is/andica. Amsterdam: H. J. Paris {Reprinted in this volume}.
DeCamp, D., 1977, "The Development of Pidgin and Creole Languages" in A. Valdman, ed. pp. 3-20.
Einarsson, S., 1949, Icelandic: Grammar, Texts, Glossary. Baltimore: John Hopkins Press.
Ferguson, C., 1977, "Baby Talk as a Simplified Register" in Talking to' Children. Ed. Snow and
Ferguson, pp. 209-235.
- - - - , and C. DeBose., 1977, uSimplified Registers, Broken Language and Pidginization" in
A. Valdman, ed. pp. 99-128.
Giv6n, T., 1979, "Prolegomena to Any Sane Creology" in Readings in Creole Studies. ed. 1. A. Han-
cock. Ghent (Belgium): Story-Scientia, pp. 3-36.
Greenberg, J. H., 1963, "Some Uhiversals of Grammar with Particular Reference to the Order of
Meaningful Elements" in Universals of Language. ed. J. H. Greenberg. Cambridge (Mass.):
MIT Press. 2nd ed. 1966, pp. 73-113.
Hall, R. Pidgin and Creole Languages. Ithaca: Cornell Univ.
Hancock, 1., 1977, "Repertoire of Pidgin and Creole Languages" in Valdman, ed., pp. 295-312.
Lafitte, P., 1962, Grammaire Basque (Navarro-Lahourdin Litteraire). Edition revue et corrigee. Re-
printed 1979, Donostia: Elkar.
Mitxelena, K., 1981, "Galdegaia eta Mintzagaia Euskaraz" in Euskal Linguistika eta Literatura:
Bide Berriak. Bilbao: U niv. de Deusto. [Reprinted in SHLV]
Rotaetxe, K., 1978, Estudio estructural del euskara de Ondarroa. Durango (Vizcaya): Leopoldo Zugaza.
Slobin, D., 1979, Psycholinguistics, 2nd ed. Glenview, IL: Scott, Foresman and Co.
Valdman, A., ed., 1977, Pidgin and Creole Linguistics. Bloomington: Indiana Univ. Press.
Whinnom, K., 1977, "Lingua Franca: I-listorical Problems" in A. Valdman, ed., pp. 295-312.